When designing rent control regulations, policy makers aim to create regulations that ensure affordable and stable housing for current tenants while minimizing exits from the rental market by landlords. Vacancy decontrol provisions that allow rent re- sets between tenants intend to strike a balance between a lower rent burden for current tenants and future potential profitability for landlords. However, such provisions also increase the incentive for landlords to evict tenants. Such evictions reduce both the anti-displacement and rent reduction effects of rent control. To study the effects of rent control on eviction behavior, we exploit variation across ZIP codes in policy exposure to the passage of the 1994 rent control referendum in San Francisco. We find that a ZIP code with the average level of treatment experiences an additional 34 eviction notices—an 83% increase—and an additional 13 wrongful eviction claims—a 125% increase. These effects were concentrated in low-income ZIP codes and were larger in years when average rent prices rose faster than the allowed rent increases for controlled units.

Policy advocates claim that one benefit of rent control may be decreased intimate partner violence (IPV). However, the theoretical effects of rent control on IPV are ambiguous. Rent control may lessen financial stressors within a relationship and decrease strain that leads to violence. However, it may make leaving the relationship more costly, shifting the bargaining power in the relationship and leading to more violence. We leverage the 1994 expansion of rent control in San Francisco as a natural experiment to study this question. This expansion created variation across zip codes in the number of rental units that were newly rent controlled. We exploit this variation in a continuous difference-in-difference design. We estimate an elasticity of -0.08 between the number of newly rent controlled units and assaults on women resulting in hospitalization. This effect translates to a nearly 10 percent decrease in assaults on women for the average zip code. This relationship is not explained by changes in neighborhood composition or overall crime, consistent with the effects being driven by individual level changes in IPV.

We examine how health insurance expansions affect the entry and location decisions of health care clinics. Exploiting county-level changes in insurance coverage following the Affordable Care Act and 1,721 retail clinic entries and exits, we find that local increases in insurance coverage do not lead to growth in the concentration of clinics on average using two-way fixed effects and instrumental variable designs. However, this null effect masks important heterogeneity by insurance type. Growth in private insurance leads to large growth in clinic entry, whereas clinic penetration is dampened by increases in Medicaid coverage. Consistent with a model in which firms face demand from both markets with administered and market-based pricing, we find that the positive (negative) supply-side effects of private insurance (Medicaid coverage) are concentrated in states with low provider reimbursements under Medicaid. We further show that similar location patterns are observed among other types of health care clinics, including urgent care centers. While it has long been accepted that reductions in the prices paid by consumers following insurance expansions should lead the supply side to expand to meet increased demand (Arrow 1963), our results demonstrate that whether health insurance expansions cause the supply side to expand or contract further depends on how the prices received by providers are affected.

Regulators often enact price restrictions with the goal of improving access to affordable products. However, the design of these regulations may interact with firm strategic entry and exit decisions in ways that mitigate the effects of pricing regulation or eliminate access to certain products entirely. In the US individual health insurance market, the Affordable Care Act established community rating areas made up of groups of counties in which insurers must offer plans at uniform prices, but insurers do not have to enter all counties in a rating area. The exact design of each market has been left to individual states. Allowing partial entry creates trade-offs in rating area design. Larger areas may support more competition, but heterogeneous areas may promote partial entry as firms choose to not enter high cost areas. To evaluate these trade-offs, I develop a model of insurer entry and pricing decisions and investigate how insurers respond to rating area design. I find that banning partial entry increases overall entry, average prices, and consumer welfare. I quantify the trade-offs of increasing rating area size and find returns to size concentrated when marginal costs are similar across counties in a rating area. Regulators must balance promoting competition with pooling high and low cost consumers in rating area design.